Pharmacy World & Science

, Volume 32, Issue 3, pp 362–372 | Cite as

Classification of drug-related problems with new prescriptions using a modified PCNE classification system

  • Patrick M. Eichenberger
  • Markus L. Lampert
  • Irene Vogel Kahmann
  • J. W. Foppe van Mil
  • Kurt E. Hersberger
Research Article

Abstract

Objectives To explore and classify drug-related problems (DRPs) with new prescriptions detected in community pharmacies using a modified PCNE (Pharmaceutical Care Network Europe) classification system. Setting Sixty-four Swiss community pharmacies offering internships for pharmacy students. Main outcome measures Occurrence, nature and pharmacist’s management of DRPs. Methods Fifth-year pharmacy students collected consecutively hospital discharge and primary care prescriptions. After training, they documented clinical and technical DRPs, causes and interventions. Results Prescriptions of 616 patients (43.0% discharged from hospital) were analysed. The patients’ median age was 56 years and they received a median of 3 (range 2–19) different drugs. In 121 (19.6%) prescriptions 141 clinical DRPs were detected. The most frequent clinical DRPs were potential drug–drug interactions (DDIs) (37.6%), drug choice (24.8%) and drug use problems (15.6%). These clinical DRPs led to a total of 299 interventions. There were 222 prescriptions (36.0%) that showed 278 technical DRPs, resulting in a total of 417 interventions. Most frequent technical DRPs were missing or unclear package size or therapy duration (32.7%) and missing or unclear dosing/application instructions (30.9%). Most DRPs (75.4%) could be managed by the pharmacist alone. The number of prescribed drugs was the main factor with an influence on the frequency of clinical and technical DRPs. Conclusion Clinical and technical DRPs are frequently observed in primary care as well as in hospital discharge prescriptions. The modified PCNE classification system, especially the amendment with a technical DRP category, proved to be useful and allowed the classification of all DRPs. Neither the setting (hospital discharge vs. primary care) nor the quality of electronically printed prescriptions, but only the number of prescribed drugs influenced the occurrence of clinical or technical DRPs.

Keywords

Classification system Community pharmacy Drug-related problems Hospital discharge PCNE Pharmaceutical care Primary care Switzerland 

Impact of findings on practice

  • In Switzerland, half of all new prescriptions show a drug-related problem; two-thirds of those are technical drug-related problems.

  • The PCNE classification V5.1 needs to be amended, to include technical drug-related problems.

  • The occurrence of clinical or technical drug-related problems is only influenced by the number of prescribed drugs. It is not relevant if the prescriptions were electronically printed.

Introduction

Many studies have shown drug-related problems (DRPs) to be very common in primary care and in hospital settings [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]. In both settings there is evidence that pharmacists’ interventions can reduce the occurrence of DRPs [6, 9, 14, 15, 17].

Receiving a newly prescribed drug is most likely an extraordinary situation for a patient who was recently informed about a diagnosis or at least was confronted with a new drug in his regimen. Because the risk of DRPs may be increased on the initiation of new drug treatments or changes within an established drug-treatment plan, a thorough consultation with a pharmacist is required to consider the new medication and, in particular, to make an accurate check for DRPs to achieve desired health outcomes [4, 18]. Thus, patients with at least one newly prescribed drug represent a relevant population to study DRPs and especially for studying the applicability of a comprehensive classification system which includes technical DRPs.

Usually, patients discharged from hospital also have new drugs prescribed and are at increased risk of experiencing DRPs [4]. Therefore, we decided to focus on both prescriptions from hospital discharge and new primary care prescriptions, expecting that there are differences between these two settings (e.g. pattern and number of DRPs) as well as differences between electronically printed and handwritten prescriptions. To our knowledge, no previous study focused on new primary care and hospital discharge prescriptions processed in Swiss community pharmacies in everyday life.

In different studies, different detection rates of DRPs were found because of variations in the methods used to identify DRPs, the classification systems and the inclusion criteria. There is no accepted standard tool for classification and documentation of DRPs, for the primary care or hospital (discharge) settings. The applicability of different systems is not yet clear, and many studies conclude that further studies are needed, with the aim to provide a tool that allows a complete classification of all DRPs that arise during prescription processing in community pharmacies.

From different pilot studies we had experience in the use of the PCNE classification (PCNE Classification V 5.01) [19]. This system attributes at least four items to each observation: (a) coding for the problem itself, (b) the actual or suspected cause(s) of the problem, (c) the intervention(s) required to resolve the DRP and (d) its outcome. We recognised as a probably important deficiency the lack of possibilities to classify technical DRPs, which arise frequently from prescription processing in community pharmacies (e.g. missing or unclear specification of drug dosage form, drug dosage and dosage regimen). In a very recent study on pharmacists’ interventions during the prescription dispensing process of new prescriptions for acute respiratory tract infections, we observed a need for addition of specifications in 11.6% of all prescription items [20]. Therefore, we planned to amend the existing PCNE classification accordingly.

The first objective of this study was to explore the occurrence, nature and pharmacist’s management of drug-related problems (DRPs) detected in community pharmacies using the modified PCNE classification system in new prescriptions. The second aim was to analyse possible differences between new primary care and hospital discharge prescriptions as well as differences between electronically printed and handwritten prescriptions. The third aim was to evaluate the applicability of the modified classification system.

Methods

Setting and participants

This prospective observational study was conducted from January to April 2007 in 64 Swiss community pharmacies offering internships for fifth-year pharmacy students. Prescriptions of patients aged 18 years or older were eligible for the study if they comprised at least two prescribed drugs and if at least one drug was new (a new drug was defined as a new drug, drug form or dosage or a generic drug but not a new package size). In Switzerland not every prescription contains a new drug because the health care system allows physicians to issue repeat prescriptions and thus enable patients to get their medications up to 12 months without further visit of the physician. Prescriptions issued by primary care physicians, physicians from ambulatory care centres or outpatient clinics of hospitals were classified as ‘primary care prescriptions’; ‘hospital discharge prescription’ was defined as a prescription issued by a hospital after the patient had been admitted for least one night.

Data collection

Sixty-four fifth-year pharmacy students had the mandatory mission from the university to collect five hospital discharge and five primary care prescriptions in a consecutive way during their pharmacy internships, leading to a convenient sample of prescriptions with at least one new medication. They were free to select one day to do so within the designated study period. While serving clients as usual they only had to identify if they were in possession of a prescription with a new drug and to check for inclusion criteria regardless whether they identified any problem or not. Then they were to check for DRPs with all the information available in a standard pharmacy, including the patient’s medication history. The software used in all Swiss community pharmacies performs computerized screening and alerts the attending employee of potential drug-drug interactions (DDIs). Whether a potential DDI was documented as a DRP or not depended on the student and the pharmacist: if the DDI was seen as a problem for the patient they had to complete a classification form.

Thereafter, prescription validation was performed as usual by a pharmacist. Accordingly, the pharmacist or the student could detect a DRP or a problem that arose during patient counselling or processing the prescription. After they had included the first prescription they were obliged to collect the subsequent nine prescriptions fulfilling the inclusion criteria leading to five hospital and five primary care prescriptions. This consecutive collecting enabled the estimation of the incidence of DRPs with new prescriptions and the compilation of a randomized, convenient sample to reflect the daily life setting in Swiss community pharmacies. For each prescription at least one data sheet had to be filled out and if multiple problems were detected in the same prescription for each problem a new data sheet was used. To each problem multiple causes and interventions could be attributed. Students were instructed to document each case shortly after dispensing the prescribed drugs. We retrieved the number of prescribed drugs per prescription from copies of prescriptions and verified plausibility of students’ documentation. Students anonymised all documents before delivery to the study centre.

Students were trained during a lecture at the university. Three case studies were used to test their performance in documenting problems, causes and interventions resulting in a proportion of 91.9% of correct coding. In addition, ongoing support was assured by the study team being in contact by e-mail or phone.

Classification of DRPs

We used the PCNE classification system for DRPs (version 5.01) [19]. This system distinguishes four dimensions: problems, causes, interventions and the corresponding outcome.

The six main categories of clinical DRPs are: (P1) adverse reaction(s), (P2) drug choice problem, (P3) dosing problem, (P4) drug use problem, (P5) interactions and (P6) other. We added a seventh category (P7) ‘technical DRP’ with the aim to be able to classify and to distinguish between “clinical DRPs” (P1–P6) and ‘technical DRPs’. A technical DRP is related to prescription quality and impedes to unambiguously dispense a drug in the correct dose, dosage form and package size (e.g. unreadable prescription, missing specifications). Accordingly, we created a seventh category of causes: C7 prescription quality (e.g. unreadable prescription, missing or unclear drug form, dosage, package size or therapy length). The pre-existing categories of causes were: (C1) drug/dose selection, (C2) drug use process, (C3) information, (C4) patient/psychological, (C5) (pharmacy) logistics and (C6) other. Categories in the intervention domain remained unchanged: (I0) ‘no intervention’, (I1) at prescriber level, (I2) at patient level, (I3) at drug level, (I4) other. Unchanged outcome categories were: (O0) outcome unknown, (O1) problem solved, (O2) problem partially solved, and (O3) problem not solved.

We created a one-side data sheet with four dimensions, 20 categories and a total of 95 items (24 problem codes, 41 causes, 23 interventions and 7 outcome codes) to enable easy data collection, including patients’ age and sex, number of medications on the prescription and whether it was a primary care or a hospital discharge prescription. In total, 14 items and one main category (technical DRP) were added to the existing classification system (Table 1).
Table 1

Added items to the original PCNE classification system V5.01

Problems (P)

Causes (C)

Interventions (I)

C7 prescription quality

P4.3 wrong or improper application method of drug (e.g. tablets without break score cut into halves)

C7.1 unreadable prescription

I2.5 patient asked for further information

P4.4 wrong or improper time of application or intake

C7.2 missing or unclear drug name, though readable

I2.6 more information retrieved from patient’s history

P7.1 technical DRP

C7.3 missing or unclear drug form if multiple forms are available

I3.7 adaptation of amount of drug dispensed

 

C7.4 missing or unclear drug potency if multiple forms are available

I4.3 information retrieved from literature

 

C7.5 missing or unclear package size or therapy length

I4.4 information retrieved from toxicology or drug information centre

 

C7.6 missing or unclear dosage or application instruction

 
 

C7.7 missing prescription of necessary applications aids

 

Users’ opinion on the usability of the tool

To evaluate usability of and satisfaction with the modified classification system, an 8-item questionnaire used in a prior published study [21] was administered to the students who used this tool. They were asked about the extent to which they agree or disagree (five-point scale) with statements regarding the usability and usefulness, satisfaction and their willingness to use the tool in their practice in the future.

Statistical analysis

All returned data sheets were processed with the automated forms processing software TeleForm® version 7.0 from Cardiff Software Inc., Vista, USA. To avoid potential errors, all numeric and letter recognitions were verified visually on data sheets and on screen.

Results are expressed as proportions and as medians with the corresponding interquartile range (IQR). The non-parametric Mann–Whitney test was used for unpaired two-sample comparisons, if data was normal distributed with Student’s t test. Statistical significance was defined as a P value <0.05. To check for possible risk factors, McNemar’s chi-square test was used. To investigate factors with influence on problems, different methods were tried, such as univariate analysis of variance, analysis of discriminance, logistic regression and correlation (bivariate, Spearman). We cite only results that proved robust in different methods of analysis. To consider the difference of prescribed number of drugs in primary care and hospital prescriptions, we analyzed influences of the setting only for a subgroup of prescriptions with ≤5 prescribed drugs. This subgroup was chosen because our sample contained few primary care prescriptions with >5 drugs and because the relationship between the number of prescribed drugs and the error rate changes significantly for >5 drugs (Fig. 1). Statistical analyses were performed using SPSS for Windows version 15.0 (SPSS Inc., Chicago, USA).
Fig. 1

Probability for prescriptions to present ‘any’ problems (n = 329), clinical (n = 121) or technical DRPs (n = 222) according to the number of prescribed drugs

Results

During the study period (January–April 2007) 625 prescriptions were collected. Nine cases did not fulfil the inclusion criteria (in one case only one drug was prescribed, eight cases had an age below 18), resulting in a sample of 616 prescriptions (57.0% primary care) equal to 616 patients receiving a total of 2,309 prescribed drugs. Table 2 shows basic study characteristics.
Table 2

Basic study characteristics

 

Primary care

Hospital discharge

Total

P-values

Prescriptions (=patients)—n (%)

351 (57.0)

265 (43.0)

616 (100.0)

Age—mean years (SD, median) (IQR, range)

53.3 (19.0;53) (28.5; 18–98)

59.3 (18.6; 62) (27.3; 19–94)

55.9 (19.0; 56) (32; 18-98)

<0.001a

Female—n (%)

212 (60.4)

139 (52.5)

351 (57.0)

0.069b

Electronic prescriptions—n (% of prescriptions)

36 (10.3)

143 (54.0)

179 (29.1)

<0.001b

Total number of prescribed drugs—n (%)

962 (41.7)

1347 (58.3)

2309

<0.001b

Number of drugs per prescription—median (IQR, range)

2 (1; 2–10)

4 (5; 2–19)

3 (2; 2–19)

<0.001a

Clinical DRPs—n (% of prescriptions)

55 (15.7)

86 (32.5)

141 (22.9)

<0.001b

Technical DRPs—n (% of prescriptions)

175 (49.9)

103 (38.9)

278 (45.1)

0.007b

Prescriptions with clinical DRPs—n (% of prescriptions)

52 (14.8)

69 (26.0)

121 (19.6)

0.001b

Prescriptions with technical DRPs—n (% of prescriptions)

135 (38.5)

87 (32.8)

222 (36.0)

0.149b

Prescriptions without any problem—n (% of prescriptions)

167 (47.6)

120 (45.3)

287 (46.6)

0.572b

aP-values for comparisons with t test (2-tailed, 95%)

bP-values for comparisons with chi-square (2-tailed, 95%)

Drug-related problems

During prescription processing in the community pharmacies, 419 clinical and technical DRPs (or both) were detected in 329 prescriptions. In relation to the total number of prescribed drugs (n = 2309) we found 6.1% (n = 141) clinical and 12.0% (n = 278) technical DRPs. All problems could be classified with the modified PCNE classification system.

Multiple clinical problems were present in 12 prescriptions, multiple technical DRPs in 49 prescriptions and 14 prescriptions showed both clinical and technical DRPs.

The analysis of the clinical problems revealed that 52 (36.9% of all clinical DRPs) were potential interactions (PCNE code P5), 35 (24.8%) drug choice problems (P2) and 22 (15.6%) drug use problems (P4). The most frequent technical DRPs were missing or unclear package size or therapy length (32.7%) and missing or unclear dosing/application instruction (30.9%). The details of clinical and technical DRPs, are shown in Tables 3 and 4, respectively.
Table 3

Clinical DRPs (n = 141) in 121 prescriptions, classified according to PCNE classification V5.1

Primary domain

Code

Detailed classification

Primary care

Hospital discharge

Total

n (% of 141)

n (% of 141)

n (% of 141)

Adverse reactions

P1

Total

  

3 (2.1)

 

P1.1

Side effect suffered (non-allergic)

1 (0.7)

1 (0.7)

2 (1.4)

 

P1.2

Side effect suffered (allergic)

1 (0.7)

1 (0.7)

 

P1.3

Toxic effects suffered

Drug choice problem

P2

Total

 

 

35 (24.8)

P2.1

Inappropriate drug (not most appropriate for indication)

6 (4.3)

5 (3.5)

11 (7.8)

 

P2.2

Inappropriate drug form (not most appropriate for indication)

4 (2.8)

3 (2.1)

7 (5.0)

 

P2.3

Inappropriate duplication of therapeutic group or active ingredient

1 (0.7)

3 (2.1)

4 (2.8)

 

P2.4

Contra-indication for drug (incl. pregnancy/breast feeding)

2 (1.4)

2 (1.4)

 

P2.5

No clear indication for drug use

6 (4.3)

4 (2.8)

10 (7.1)

 

P2.6

No drug prescribed but clear indication

1 (0.7)

1 (0.7)

Dosing problem

P3

Total

 

 

18 (12.8)

 

P3.1

Drug dose too low or dosage regimen too long

2 (1.4)

2 (1.4)

4 (2.8)

 

P3.2

Drug dose too high or dosage regimen too frequent

1 (0.7)

7 (5.0)

8 (5.7)

 

P3.3

Duration of treatment too short

2 (1.4)

2 (1.4)

4 (2.8)

 

P3.4

Duration of treatment too long

1 (0.7)

1 (0.7)

2 (1.4)

Drug use problem

P4

Total

 

 

22 (15.6)

 

P4.1

Drug not taken/administered at all

 

P4.2

Wrong drug taken/administered

1 (0.7)

1 (0.7)

 

P4.3

Wrong/not appropriate drug use/application

5 (3.5)

4 (2.8)

9 (6.4)

 

P4.4

Wrong/not appropriate time of use/application

8 (5.7)

4 (2.8)

12 (8.5)

Interactions

P5

Total

 

 

53 (37.6)

 

P5.1

Potential interaction

10 (7.1)

42 (29.8)

52 (36.9)

 

P5.2

Manifest interaction

1 (0.7)

1 (0.7)

Others

P6

Total

 

 

10 (7.1)

 

P6.1

Patient dissatisfied with therapy despite taking drug(s)

 

P6.2

Insufficient awareness of health and diseases (possibly leading to future problems)

4 (2.8)

6 (4.3)

10 (7.1)

 

P6.3

Unclear complaints, further clarification necessary

 

P6.4

Therapy failure (reason unknown)

Total

  

55 (39.0)

86 (61.0)

141 (100.0)

Table 4

Technical DRPs (n = 278) in 222 prescriptions

Technical DRPsa

Primary care

Hospital discharge

Total

n (% of 278)

n (% of 278)

n (% of 278)

Missing prescription of necessary application aids

1 (0.4)

2 (0.7)

3 (1.1)

Missing/unclear drug name though legible

6 (2.2)

8 (2.9)

14 (5.0)

Missing/unclear drug form, if several available

17 (6.1)

6 (2.2)

23 (8.3)

Missing/unclear drug potency, if several available

15 (5.4)

13 (4.7)

28 (10.1)

Unreadable prescription

26 (9.4)

7 (2.5)

33 (11.9)

Missing/unclear dosing/application instruction

64 (23.0)

22 (7.9)

86 (30.9)

Missing/unclear package size and/or therapy length

46 (16.5)

45 (16.2)

91 (32.7)

Total

175 (62.9)

103 (37.1)

278 (100.0)

a≥1 technical DRP per prescription possible

The main contributing factor for presentation of clinical and technical DRPs was the number of drugs per prescription. Other variables had no significant influence in our dataset (univariate analysis of variance, analysis of discriminance, logistic regression and correlation).

The probability for a clinical DRP was similar in prescriptions between two and five drugs; if >5 drugs were prescribed then the probability increased (see Fig. 1). The number of drugs is correlated with the chance of a clinical DRP (r = 0.220; P < 0.001; bivariate correlation, Spearman) and with ‘any problems’ (r = 0.132; P = 0.001; bivariate correlation, Spearman) but not with technical DRPs (r = 0.012; P = 0.775; bivariate correlation, Spearman).

The mean number of prescribed drugs was significantly higher for prescriptions from hospital discharge than from primary care (5.1 vs. 2.7; P < 0.001; t test). This higher number of prescribed drugs in hospital discharge prescriptions caused more clinical but not more technical DRPs. To analyse the influence of the setting (primary care and hospital discharge), we only considered prescriptions with ≤5 prescribed drugs because primary care prescriptions rarely included >5 prescribed drugs. We observed significantly more technical DRPs among prescriptions from primary care (P = 0.043; chi-square; 95%; 2-tailed) but no influence of the setting for any problem (P = 0.080; chi-square; 95%; 2-tailed) or clinical DRPs (P = 0.924; chi-square; 95%; 2-tailed). There was no association between gender and problems. Because the relationship between the number of prescribed drugs and technical DRPs is not linear and we do not have sufficient data we could not statistically correct for this effect.

Causes

To each problem a maximum of three causes could be attributed. Overall, 166 causes for clinical DRPs were reported, with a majority (50.0%, n = 83) that was related to the selection of the drug and/or dosage schedule (C1). The second and third most common causes with 15.7% (n = 26) involved the information process (C3) and the drug use process with 14.5% (n = 24; C2; see Table 5).
Table 5

Top 10 of the most frequently reported causes (n = 166) induced by clinical DRPs (n = 141), classified according to PCNE classification V5.1

Primary domain

Code

Detailed classification

Primary care

Hospital discharge

Total

n (% of 166)

n (% of 166)

n (% of 166)

Drug/dose selection

C1

Total

 

 

83 (50.0)

C1.1

Inappropriate drug selection

17 (10.2)

38 (22.9)

55 (33.1)

 

C1.2

Inappropriate dosage selection

5 (3.0)

9 (5.4)

14 (8.4)

 

C1.4

Pharmacokinetic problems, incl. ageing/deterioration in organ function and interaction

3 (1.8)

7 (4.2)

10 (6.0)

Drug use process

C2

Total

 

 

24 (14.5)

C2.1

Inappropriate timing of administration and/or dosing intervals

5 (3.0)

8 (4.8)

13 (7.8)

 

C2.3

Drug overused/over-administered

2 (1.2)

3 (1.8)

5 (3.0)

 

C2.6

Patient unable to use drug (form) as directed

3 (1.8)

1 (0.6)

4 (2.4)

Information

C3

Total

  

26 (15.7)

 

C3.1

Instructions for use/taking not known

9 (5.4)

6 (3.6)

15 (9.0)

 

C3.2

Patient unaware of reason for drug treatment

4 (2.4)

4 (2.4)

Patient/Psychological

C4

Total

  

14 (8.4)

 

C4.8

Burden of therapy

2 (1.2)

3 (1.8)

5 (3.0)

(Pharmacy) Logistics

C5

Total

 

 

5 (3.0)

C5.1

Prescribed drug not available

1 (0.6)

3 (1.8)

4 (2.4)

Interventions

A total of 716 interventions (see Table 6) were reported for all prescriptions (1.2 per 616 prescriptions/2.2 per 329 prescriptions with problems). All 141 clinical DRPs induced 299 (41.8% out of 716) interventions (2.1 per clinical DRP/2.5 per prescription with clinical DRPs/0.5 per 616 prescriptions; range 1–7) and 278 technical DRPs induced 417 (58.2% out of 716) interventions (1.5 per technical DRP/1.9 per prescription with a technical DRP/0.7 per 616 prescriptions; range of 0–6).
Table 6

Top 10 of the most frequently reported interventions (n = 716) induced by clinical or technical DRPs, classified according to PCNE classification V5.1

Primary domain

Code

Detailed classification

Primary care

Hospital discharge

Total

   

Clinical DRPs

Technical DRPs

Clinical DRPs

Technical DRPs

All DRPs

   

n (% of 716)

n (% of 716)

n (% of 716)

n (% of 716)

n (% of 716)

At prescriber level

I1

Total

   

 

102 (14.2)

I1.1

Prescriber informed only

2 (0.3)

3 (0.4)

7 (1.0)

2 (0.3)

14 (2.0)

I1.2

Asked prescriber for further information

7 (1.0)

26 (3.6)

14 (2.0)

6 (0.8)

53 (7.4)

I1.3

Intervention proposed, approved by prescriber

6 (0.8)

11 (1.5)

8 (1.1)

3 (0.5)

28 (3.9)

At patient/carer level

I2

Total

 

 

 

 

171 (23.9)

I2.1

Patient (medication) counselling

25 (3.5)

66 (9.2)

63 (8.8)

42 (5.9)

196 (27.4)

I2.5

Patient asked for further information

21 (2.9)

83 (11.6)

25 (3.5)

28 (3.9)

157 (21.9)

I2.6

Further information retrieved from patient history

4 (0.6)

20 (2.8)

21 (2.9)

9 (1.3)

54 (7.5)

At drug level

I3

Total

 

 

 

 

94 (13.1)

I3.1

Drug changed

4 (0.6)

9 (1.3)

5 (0.7)

3 (0.4)

21 (2.9)

I3.2

Dosage changed

4 (0.6)

4 (0.6)

4 (0.6)

1 (0.1)

13 (1.8)

I3.4

Instructions for use changed

9 (1.3)

9 (1.3)

3 (0.4)

3 (0.4)

24 (3.4)

I3.7

Amount of drug to dispense changed

5 (0.7)

5 (0.7)

4 (0.6)

4 (0.6)

18 (2.5)

To manage all problems a total of 81 (13.1% out of 616 prescriptions) direct contacts to the prescriber (call/conversation) were necessary: 42 (6.8% out of 616) prescriptions with one or more clinical DRPs and 46 (7.5% out of 616) prescriptions with one or more technical DRPs led to a direct contact with the prescriber (seven prescriptions with both a clinical DRP and a technical DRP). One hundred and seventy-five (63.0% of 278) of technical DRPs could be solved by the pharmacy together with the patient, but only 56 (39.7% out of 141) of clinical DRPs were solved in cooperation with the patient.

The 299 interventions induced by clinical DRPs involved 178 drugs (in 48 cases two drugs were related to one problem, e.g. interaction). Cardiovascular drugs (C; 21.9%, n = 39) were most often involved, followed by nervous system (N; 21.3%, n = 38) and musculo-skeletal system drugs (M; 10.1%, n = 18).

Users’ opinion

After training and data collection, observers were asked eight questions about the acceptability and usability of the classification system they had used in this study (see Table 7). Most of the users think that it is important to have an opportunity to document the efforts of the pharmacies but only one-third agreed that the tool is easy to use and practical.
Table 7

Users’ opinion (n = 64) of the tool

 

Mean score ± SD

Agree or strongly agree

Neutral

Disagree or strongly disagree

  

n (%)

n (%)

n (%)

1. The classification system was comprehensive and included all drug-related problems I identified.

3.4 ± 1.1

29 (45.3)

18 (28.1)

17 (26.6)a

2. I did not have problems finding out the proper classification of drug-related problems I identified.

2.9 ± 1.0

18 (28.1)

23 (35.9)

23 (35.9)a

3. The classification system was easy to use and practical.

3.1 ± 0.9

22 (34.4)

28 (43.8)

14 (21.9)a

4. I will use the classification system in my practice in the future.

2.3 ± 1.3

5 (7.8)

14 (21.9)

41 (64.1)b

5. In general I am satisfied with the classification system.

3.3 ± 1.0

30 (46.9)

19 (29.7)

15 (23.4)a

6. The expenditure of time to classify the problems was adequate.

3.7 ± 1.0

42 (65.6)

12 (18.8)

10 (15.6)a

7. I think it is important to have an opportunity to classify the effort of pharmacies.

4.4 ± 0.7

56 (87.5)

7 (10.9)

8. The PCNE classification would be a good tool to document the activities of pharmacies.

3.1 ± 1.1

21 (32.8)

22 (34.4)

19 (29.7)a

1 strongly disagree, 2 disagree, 3 neutral, 4 agree, 5 strongly agree

aNo user answered this question with ‘strongly disagree’

bTwenty users answered this question with ‘strongly disagree’

Discussion

This study examined the frequency, nature and pharmacist’s management of DRPs with primary care and hospital discharge prescriptions which comprised at least one new drug. We found a high occurrence of clinical (19.6%) and, in particular, of technical DRPs (36.0%). More than 50% of all prescriptions showed a clinical or a technical DRP or both. Compared to other studies our numbers are quite high but in this study we set out to use prescriptions which we considered likely to have a high prevalence, i.e. newly started drugs and prescriptions with at least two drugs. In 2007 Hammerlein found only 0.93 DRPs per 100 patients and 1.16 DRPs per 100 prescriptions [5]. Westerlund conducted a survey on DRPs in prescription-only medicines in 1999 and found 2.8 DRPs per 100 patient contacts [1]. A recent Swiss study detected 0.74 clinical DRPs per 100 prescriptions with the need of an intervention and 1.9 technical DRPs per 100 prescriptions [8]. These figures illustrate the problem that such comparisons are hampered by different settings, measurements methods and classification systems. Paulino [4] found in a study in six European countries even more DRPs than we did (103.7 DRPs per 100 patients discharged from hospital; 63.7% out of all patients had at least one DRP). Uncertainty or lack of knowledge about the aim of the drug (29.5%), side effects (23.3%) and practical problems (7.6%) were the most common DRPs. However, only few DDIs (4.0%) were detected, probably because most of them did not have access to patient medication histories. In Switzerland every pharmacy is equipped with software that includes an automatic DDI screening system and participants in our study probably indicated every possible potential interaction, even if not clinical relevant. Therefore, the most frequently reported clinical DRPs in our study were DDIs (37.6% of all clinical DRPs; 7.0% of all patients). In the German study [5] DDIs were also most frequent but the fraction was only 8.5% of all DRPs. In another primary care study DDIs were the third most common problem with 3.2%, and in a study performed in hospital setting they observed 17.0% DDIs [8, 9]. Paulino et al. found 60.0% of patients after hospital discharge with a potential interaction [4].

However, the comparison of the study’s findings with those of other studies is difficult due to differences in setting and aim [1, 4, 5, 8, 9, 22]. Furthermore, the frequency of detected problems can be influenced by the systematic screening under study conditions, which differs significantly from daily practice (Hawthorne effect [23]). In our study, students and pharmacists probably had a more positive attitude towards the provision of pharmaceutical care (e.g. check for DRPs) and they probably detected most possible problems. This may explain our high detection rate of 22.9. If DDIs are as frequent as found in this or other studies more emphasis should be placed on best management of potential DDIs in community pharmacies. However, if we omit all interactions we still found a detection rate of 14.3.

Using fifth-year pharmacy students for data collection during their internship and collecting primary care as well as hospital discharge prescriptions in a continuous way, enabled us to compile a randomized, convenient sample and to estimate the prevalence of DRPs with new prescriptions. We had instructed the students to perform prescription processing as usual. Thus, every prescription was checked by a pharmacist before dispensing and we were not depending on inferior competencies of students. Factors associated with clinical and technical DRPs were tested with several statistical methods and the main factor influencing the presence of any clinical problems was the number of drugs. Older people showed both more clinical and technical DRPs, but this was to be expected because the number of drugs increased with age. In addition, we expected the setting to have an influence on the occurrence of DRPs but different statistical tests did not show any relationship. More clinical DRPs were found in hospital discharge than in primary care prescriptions but the mean number of prescribed drugs was higher on these prescriptions, which explains the higher frequency of problems. Surprisingly, more technical DRPs were found in primary care prescriptions with fewer prescribed drugs than in hospital discharge prescriptions.

Most problems were caused by drug selection and/or drug dose selection and many problems were solved after discussion with patients or family members. More than two-thirds of technical but less than half the clinical DRPs could be solved by the pharmacy together with the patient. Pharmacists should therefore actively involve patients and relatives in the screening and solving process for DRPs. Almost two-thirds of all interventions and 50% of all contacts with the prescriber were related to technical DRPs which shows their importance and frequency in daily life practice in community pharmacies. Therefore, inclusion of technical DRPs into the classification is strongly recommended. Looking at the intervention part, pharmacists indicate the importance of the detected clinical DRPs, in particular activities at prescriber and drug level. The high frequency of interventions is related to the high detection rates of DRPs.

We amended in this study the original PCNE system with 80 items by addition of 15 items (incl. technical DRPs as one new main category) resulting in a total of 95 items. This new problem category was specified with seven items to receive enough information. As suggested by Lampert [9] and Allenet [24], we further added two items in the category ‘drug use problem’ to categorise wrong or improper use of a drug and wrong or improper time of applying a drug. A previous study showed a documentation rate of 97.8% in a hospital setting without any modification of the classification system but they reported a lack of certain items for in-patients [9]. Therefore, an important result of this study is the fact that our modified system allowed a complete classification of all problems. In contrast, a study [5] with the PI-Doc system with 72 items was amended with 27 items but 362 out of 10,427 (3.5%) cases still could not be classified. Our study found twice as many technical DRPs as clinical ones. Most of the documentation systems do not separate technical from clinical DRPs and give them the same level of importance, or they have integrated technical DRPs in ‘prescribing errors’ or ‘other’ [1, 19, 25, 26, 27]. Often there is only a definition for clinical but not for technical DRPs. The German study of 2007 reported 8.3% (out of 10,427) of technical DRPs and difficulties to classify technical DRPs in over 50% of the cases [5]. Our results strongly support the inclusion of a new problem category ‘technical DRPs’ with enough further possibilities to specify them. Furthermore, different classification systems should feature at least the same main categories with the possibility to change single items to be able to compare results from different studies.

Our study has strengths and limitations. We selected prescriptions from primary care during prescription processing in a daily life setting. We could ascertain the reported problems with the copies of the prescriptions and we could retrospectively differentiate between handwritten and electronically written prescriptions. All problems could be coded with our modified PCNE classification system. However, we did not assess the level of severity and the clinical or economic impact of clinical and technical DRPs. Furthermore, we cannot assure that the prescriptions selected were consecutive ones and the diversity of persons who documented clinical and technical DRPs could lead to variability in coding for the same problem, even though the participant observers were trained with specific and detailed instructions. In our study the PCNE system showed to be easy to use for a majority of the users and took little time, as it has been created for the documentation of DRPs in the community pharmacy setting. However, one quarter of students evaluated the PCNE tool as not comprehensive although they had classified all problems. Our students rated very critically all questions if comparing with the study from which the questions were retrieved [21].

Our study reflects only the situation in Switzerland. A recent multinational study by Mc Elnay et al. investigated the extent to which pharmaceutical care has already been implemented into daily community pharmacy practice across Europe (Mc Elnay J, Hughes C. Provision of pharmaceutical care by community pharmacists: a comparison across Europe). The behavioural pharmaceutical care scale was used and the mean of all 13 countries was 72.1 ± 8.5 (mean ± SD) with a range of mean scores from 50.6 to 83.5 (possible score: 15–160) reflecting a considerable variation between countries. Switzerland had a score of 73.2 ± 18.7 (mean ± SD) indicating that Swiss pharmacists behave most probably not very differently from their counterparts in other European countries. Furthermore, this study revealed an important lack of documentation activities throughout Europe: only 25.2% of all possible documentation possibilities were used. Thus, a documentation and classification system is urgently needed. Intensive training can increase the number of interventions without increasing the time spent on documentation [28, 29]. As Lampert [9] reported, the PCNE system with its classification of each DRP on four different levels gives enough detail to allow qualitative and even economic analyses.

Conclusion

In the delivery process of new prescribed medications, clinical and technical DRPs are frequently observed in new primary care and in hospital discharge prescriptions. Neither the setting (hospital discharge vs. primary care) nor the quality of electronically printed prescriptions, but only the number of prescribed drugs influenced the occurrence of clinical or technical DRPs in this study. Most DRPs could be managed by the pharmacist alone or after discussion with the patient but without any contact to the prescriber. Therefore, the management of DRPs in community pharmacies is a very important activity which should be explored more intensively in further studies. The modified PCNE classification system, especially the amendment with a technical DRP category, proved to be useful and allowed the classification of all DRPs, but still rather complicated to apply in pharmacy practice.

Notes

Acknowledgements

We would like to thank the owners of the 64 participating pharmacies and Michael Mittag for the support in statistical analysis.

Funding

No funding was provided for this Ph.D. student research.

Conflicts of Interest

None.

References

  1. 1.
    Westerlund LT, Almarsdottir AB, Melander A. Drug-related problems and pharmacy interventions in community practice. Int J Pharm Pract. 1999;7(11):40–50.Google Scholar
  2. 2.
    Ernst ME, Iyer SS, Doucette WR. Drug-related problems and quality of life in arthritis and low back pain sufferers. Value Health. 2003;6(1):51–8.CrossRefPubMedGoogle Scholar
  3. 3.
    Triller DM, Clause SL, Briceland LL, Hamilton RA. Resolution of drug-related problems in home care patients through a pharmacy referral service. Am J Health Syst Pharm. 2003;60(9):905–10.PubMedGoogle Scholar
  4. 4.
    Paulino EI, Bouvy ML, Gastelurrutia MA, Guerreiro M, Buurma H. Drug related problems identified by European community pharmacists in patients discharged from hospital. Pharm World Sci. 2004;26(6):353–60.PubMedGoogle Scholar
  5. 5.
    Hammerlein A, Griese N, Schulz M. Survey of drug-related problems identified by community pharmacies. Ann Pharmacother. 2007;41(11):1825–32.CrossRefPubMedGoogle Scholar
  6. 6.
    Krahenbuhl-Melcher A, Schlienger R, Lampert M, Haschke M, Drewe J, Krahenbuhl S. Drug-related problems in hospitals: a review of the recent literature. Drug Saf. 2007;30(5):379–407.CrossRefPubMedGoogle Scholar
  7. 7.
    Howard R, Avery A, Bissell P. Causes of preventable drug-related hospital admissions: a qualitative study. Qual Saf Health Care. 2008;17(2):109–16.CrossRefPubMedGoogle Scholar
  8. 8.
    Krahenbuhl JM, Kremer B, Guignard B, Bugnon O. Practical evaluation of the drug-related problem management process in Swiss community pharmacies. Pharm World Sci. 2008;30(6):777–86.CrossRefPubMedGoogle Scholar
  9. 9.
    Lampert ML, Kraehenbuehl S, Hug BL. Drug-related problems: evaluation of a classification system in the daily practice of a Swiss University Hospital. Pharm World Sci. 2008;30(6):768–76.CrossRefPubMedGoogle Scholar
  10. 10.
    Patel P, Zed PJ. Drug-related visits to the emergency department: how big is the problem? Pharmacotherapy. 2002;22(7):915–23.CrossRefPubMedGoogle Scholar
  11. 11.
    Becker ML, Kallewaard M, Caspers PW, Visser LE, Leufkens HG, Stricker BH. Hospitalisations and emergency department visits due to drug-drug interactions: a literature review. Pharmacoepidemiol Drug Saf. 2007;16(6):641–51.CrossRefPubMedGoogle Scholar
  12. 12.
    Rupp MT, DeYoung M, Schondelmeyer SW. Prescribing problems and pharmacist interventions in community practice. Med Care. 1992;30(10):926–40.CrossRefPubMedGoogle Scholar
  13. 13.
    Smith CP, Christensen DB. Identification and clarification of drug therapy problems by Indian health service pharmacists. Ann Pharmacother. 1996;30(2):119–24.PubMedGoogle Scholar
  14. 14.
    Blix HS, Viktil KK, Reikvam A, Moger TA, Hjemaas BJ, Pretsch P, et al. The majority of hospitalised patients have drug-related problems: results from a prospective study in general hospitals. Eur J Clin Pharmacol. 2004;60(9):651–8.CrossRefPubMedGoogle Scholar
  15. 15.
    Viktil KK, Blix HS, Moger TA, Reikvam A. Interview of patients by pharmacists contributes significantly to the identification of drug-related problems (DRPs). Pharmacoepidemiol Drug Saf. 2006;15(9):667–74.CrossRefPubMedGoogle Scholar
  16. 16.
    Shalansky S, Nakagawa R, Wee A. Drug-related problems identified and resolved using pharmaceutical care versus traditional clinical monitoring. Can J Hosp Pharm. 1996;49(6):7.Google Scholar
  17. 17.
    Hanlon JT, Lindblad CI, Gray SL. Can clinical pharmacy services have a positive impact on drug-related problems and health outcomes in community-based older adults? Am J Geriatr Pharmacother. 2004;2(1):3–13.CrossRefPubMedGoogle Scholar
  18. 18.
    Barber N, Parsons J, Clifford S, Darracott R, Horne R. Patients’ problems with new medication for chronic conditions. Qual Saf Health Care. 2004;13(3):172–5.CrossRefPubMedGoogle Scholar
  19. 19.
    PCNE. The PCNE Classification V 5.01 2006 29 May 2006 [cited 20 Sep 2009]. Available from: http://www.pcne.org/dokumenter/DRP/PCNE%20classification%20V5.01.pdf.
  20. 20.
    Hersberger K, Botomino A, Sarkar R, Tschudi P, Bucher H, Briel M. Prescribed medications and pharmacy interventions for acute respiratory tract infections in Swiss primary care. J Clin Pharm Ther. 2009;34(4):387–95.CrossRefPubMedGoogle Scholar
  21. 21.
    AbuRuz SM, Bulatova NR, Yousef AM. Validation of a comprehensive classification tool for treatment-related problems. Pharm World Sci. 2006;28(4):222–32.CrossRefPubMedGoogle Scholar
  22. 22.
    Westerlund LT, Marklund BR, Handl WH, Thunberg ME, Allebeck P. Nonprescription drug-related problems and pharmacy interventions. Ann Pharmacother. 2001;35(11):1343–9.CrossRefPubMedGoogle Scholar
  23. 23.
    Eccles M, Grimshaw J, Campbell M, Ramsay C. Research designs for studies evaluating the effectiveness of change and improvement strategies. Qual Saf Health Care. 2003;12(1):47–52.CrossRefPubMedGoogle Scholar
  24. 24.
    Allenet B, Bedouch P, Rose FX, Escofier L, Roubille R, Charpiat B, et al. Validation of an instrument for the documentation of clinical pharmacists’ interventions. Pharm World Sci. 2006;28(4):181–8.CrossRefPubMedGoogle Scholar
  25. 25.
    Knapp KK, Katzman H, Hambright JS, Albrant DH. Community pharmacist interventions in a capitated pharmacy benefit contract. Am J Health Syst Pharm. 1998;55(11):1141–5.PubMedGoogle Scholar
  26. 26.
    Leemans L, Veroeveren L, Bulens J, Hendrickx C, Keyenberg W, Niesten F, et al. Frequency and trends of interventions of prescriptions in Flemish community pharmacies. Pharm World Sci. 2003;25(2):65–9.CrossRefPubMedGoogle Scholar
  27. 27.
    Schaefer M. Discussing basic principles for a coding system of drug-related problems: the case of PI-Doc. Pharm World Sci. 2002;24(4):120–7.CrossRefPubMedGoogle Scholar
  28. 28.
    Benrimoj SI, Langford JH, Berry G, Collins D, Lauchlan R, Stewart K, et al. Economic impact of increased clinical intervention rates in community pharmacy. A randomised trial of the effect of education and a professional allowance. Pharmacoeconomics. 2000;18(5):459–68.CrossRefPubMedGoogle Scholar
  29. 29.
    Currie JD, Chrischilles EA, Kuehl AK, Buser RA. Effect of a training program on community pharmacists’ detection of and intervention in drug-related problems. J Am Pharm Assoc (Wash). 1997;NS37(2):182–91.Google Scholar

Copyright information

© Springer Science+Business Media B.V. 2010

Authors and Affiliations

  • Patrick M. Eichenberger
    • 1
  • Markus L. Lampert
    • 2
  • Irene Vogel Kahmann
    • 3
  • J. W. Foppe van Mil
    • 4
  • Kurt E. Hersberger
    • 1
  1. 1.Pharmaceutical Care Research GroupUniversity of BaselBaselSwitzerland
  2. 2.Clinical and Hospital PharmacyCantonal Hospital BruderholzBruderholzSwitzerland
  3. 3.Clinical and Hospital PharmacyCantonal Hospital SchaffhausenSchaffhausenSwitzerland
  4. 4.van Mil ConsultancyZuidlarenNetherlands

Personalised recommendations